Freedom of Choice vs. High Wages – Evidence from a Gift Exchange Experiment

How to cite

Krügel, Jan Philipp; Meemann, Christine. 2022. "Freedom of Choice vs. High Wages – Evidence from a Gift Exchange Experiment." AEA RCT Registry. AEARCTR-0008785. https://doi.org/10.1257/rct.8785-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Jan Philipp Krügel* , Christine Meemann

* Primary investigator

Registry ID
AEARCTR-0008785
DOI
https://doi.org/10.1257/rct.8785-1.0
First registered
Last updated
Status
In development
Trial dates
2022-01-23 – 2022-04-30
Countries
Private
JEL codes
J30; L20; C91
Registry record
http://www.socialscienceregistry.org/trials/8785

Abstract

Using a laboratory experiment, we analyze whether freedom of choice or high wages induce higher work effort. In our gift exchange experiment, participants have to work on one of three different real effort tasks. After a trial round, participants are either assigned to a task or they are able to choose their preferred task themselves (freedom of choice). Furthermore, they either receive a normal wage or an unexpected wage increase (high wages). The wage increase equals the mean willingness-to-accept (WTA) to sell one’s decision right. The WTA is determined in a pre-study with a different group of participants. After having completed the real effort task, we ask the participants whether they are willing to work on another task that is beneficiary to the experimenter without receiving additional compensation. Given the evidence of other gift exchange experiments, we conjecture that both freedom of choice and high wages lead to an increase in voluntary work provision compared to a baseline scenario without any of these two incentives. Which of the two gifts induces more effort, however, is an open question that we want to answer.

Intervention

In a between-subjects design, we vary the following parameters: (i) participants either receive a normal wage or a high wage and (ii) participants are either assigned to a task in the first stage of the experiment (no decision right) or they are able to choose their preferred task in the first stage themselves (decision right).
We run three different treatments. Additionally, we conduct a pre-study:
- Pre-study: pre-study to determine the willingness-to-accept (WTA) to sell one's decision right to the experimenter. We use the mean WTA of participants in the pre-study as the wage increase in Treatment 2 (high wage = normal wage + WTA).
- Treatment 1 (Control): normal wage and no decision right.
- Treatment 2 (High Wage): high wage and no decision right.
- Treatment 3 (Freedom of Choice): normal wage and decision right.

Experimental design

The experiment has four parts. Participants receive sperate instructions for each part. The instructions for each part are displayed on the screen immediately before the start of the respective part.
In part 1, participants have to work on one of three different real effort tasks: slider task, encryption task or counting zeroes. The real effort tasks are similar in structure, but differ enough so that participants can form a true preference. For this reason, we also decided to implement three real effort tasks. The literature suggests that freedom of choice exists if there are at least two different options available. Nevertheless, to provide participants with a more meaningful choice, we provide three options.
First, we conduct a trial round, where participants have to solve one block of each of the three tasks. Then, they have to work on one of the tasks for 15 minutes. At this stage, our treatment variation comes into place. We vary whether subjects are assigned to a task (T1 and T2) or whether they can choose the task themselves (T3). In the pre-study, which is conducted first, we ask the participants if they want to choose the task themselves or if they want to sell their decision right to the experimenter for different amounts of money. We use an incentive-compatible choice list to elicit the WTA in the pre-study. There are several items, each with two choices: participants can either keep or sell the decision right. The amount of money obtained from the sale of the decision right increases with each item (the maximum is 5 Euros). After all decisions in the choice list are made, one item is randomly selected and the choice made in this item is implemented. If the decision right is sold to the experimenter, the participant receives the amount of money specified in the item and the participant is assigned to a task by the experimenter.
In T1, T3 and the pre-study, participants receive a flat payment of 10 Euros for stage 1 (“normal wage”). In T2, the participants receive 10 Euros and an additional unexpected wage increase, which is equal to the mean WTA to sell the decision right of participants in the pre-study. However, in all treatments, participants only receive the payment (normal wage or high wage, respectively) if they correctly solve a minimal number of items in the real effort task. If not, they do not receive a payment.
In the pre-study, we elicit the willingness-to-accept instead of the willingness-to-pay because we are interested in the value of the gift that we provide to the participants (instead of the value that the participants are willing to pay to receive the gift). Moreover, we ensure that participants in the pre-study have the choice between receiving a normal wage and a wage increase from selling their decision right (as in T2) or a normal wage and freedom of choice (as in T3).
After all participants have completed the task from part 1, we ask several non-incentivized questions about individual perceptions of the task. Additionally, we ask the participants who did not choose a task themselves which task they would have preferred to work on. In T3, we also ask participants for how much money they would have sold their decision right to the experimenter.
In part 2, participants are asked if they are willing to work on an additional task that is beneficial to the experimenter without receiving extra compensation. The task is to evaluate different real effort tasks. The real effort tasks and the evaluation criteria are taken from Charness et al. (2018). Since the participants worked on a real effort task themselves in part 1, it should be clear that the experimenters are interested in the general functionality and the outcomes of different real effort tasks. Furthermore, we explicitly tell the participants that working on the additional task is beneficial to the experimenters. This is no deception since we are interested in the perceived differences between the tasks: The results are relevant for the current research project presented here and for future projects of ours as well.
The additional task has 10 items. There is no time restriction. However, not all 10 real effort tasks have to be evaluated: the participants can stop working at any time. Once a participant decides to stop working, he goes straight to part 3. The same applies if a participant decides to not work on the additional task at all. We measure productivity in part 2 by counting how many real effort tasks have been fully evaluated.
In part 3, we conduct an incentivized cognitive ability task. The task is based on the IQ test by Frederick (2005) and consists of three questions. We use the results of the test to control for whether cognitive ability has an effect on effort.
In part 4, we conduct a charity dictator game. Participants receive 2 Euros and can make a donation to a charity of their choice (we provide them with a list of seven charities). We use the results of the dictator game to control for whether effort is provided due to reciprocity or altruism.

Design details

Randomization method
Randomization is done by a computer and through O-Tree.
Randomization unit
Individual
Clusters
50-75 participants per treatment (200-300 in total)
Observations
200-300 participants, 200-300 observations
Arms
2 sessions per treatment
Primary outcomes
We hypothesize that both higher wages and freedom of choice lead to an increase in work effort compared to a scenario without these two incentives. Hence, we expect effort to be higher in the treatments T2 (High Wage) and T3 (Freedom of Choice) compared to T1 (Control). Which of the two gifts induces more effort, however, is an open question that we want to answer. In T2, the wage increase equals the WTA to sell one's decision right. The WTA encompasses two parts: (i) the instrumental value of the decision right, i.e., the difference between the value of the most preferable option and the default option. Since there is no default option, the participants form an expectation about their subjective instrumental value. (ii), the intrinsic value of the decision right, i.e., the pure value of being able to decide and to make a decision. We have two outcome variables that measure effort (one primary and one secondary outcome variable). Primary outcome variable: First, we study whether and to what extent participants complete the additional task in each of the three treatments. The additional task is clearly beneficial to the experimenter. Therefore, receiving a gift from the experimenter (high wage or freedom of choice) may induce participants to reciprocate the gift by working on the additional task despite receiving no extra pay. In T1 (Control), participants do not receive such a gift. By setting the mean WTA to sell one’s decision right as the wage increase in T2, the effort levels in the treatments T2 (High Wage) and T3 (Freedom of Choice) are comparable. We intend to use OLS regressions with T1 as a benchmark, controlling for cognitive ability (derived from a separate cognitive ability task) and altruism (derived from giving behavior in the dictator game). If we find that T2 has a stronger influence on effort than T3, we can conclude that receiving the monetary gift of a high wage induces subjects to work harder than receiving the non-monetary gift of freedom of choice. If we find that T3 has a stronger influence on effort than T2, we can draw the opposite conclusion. When comparing T1 and T3, we have to consider that the decision right (and the WTA) encompasses the instrumental value as well as the intrinsic value. We are interested in both. However, we can also isolate the intrinsic value of the decision right by looking only at subjects who received their preferred real effort task in T1 and then comparing their mean effort to that of subjects in T3.
Secondary outcomes
Second, we compare the productivity of the participants in the real effort task in each of the three treatments. We measure productivity by counting the items that the participants successfully completed in the given time frame. Since the real effort tasks may differ in scope and difficulty, we can only compare productivity for the same real effort task across the different treatments. A potential problem with this variable (items solved) is that participants may not perceive their (extra) work as a gift to the experimenter because the real effort tasks are artificial.

Cited by

No citation data available for this registration yet.

Keywords